Macroeconomic Framework for Development in Gujarat

Transcription

1 Macroeconomic Framework for Development in Gujarat -Ravindra H. Dholakia * I.I.M. Ahmedabad rdholkia@iimahd.ernet.in ABSTRACT The paper begins by discussing need for developing a regional accounting framework and estimating relevant variables to formulate realistic growth targets and appropriate development strategy in the reform era for the Gujarat State as a case. The study then examines growth experience in various sectors of the state and derives growth potential of the economy in medium and long term. Elementary regional accounting framework and estimates of crucial macroeconomic aggregates at the state level in India are attempted for the first time for Gujarat to derive implications on resource availability and investment requirement to achieve alternative growth targets. The study finds that Gujarat is a high saving society comparable to China and Korea, but invests much less domestically. Although it a net importer internationally, it is a major net exporter within the country. The paper also attempts to identify the prime movers or principal drivers of the economic growth in the state by fitting a simultaneous equations model on the recent time series data on Gujarat. Electricity, gas & water supply; storage & communications; construction; real estates and rainfall are the prime movers in Gujarat. The paper also examines the social and human development aspects and explores how they can be integrated with the macroeconomic growth model in Gujarat. The paper concludes by discussing strategic policy interventions to achieve the development goals of the state. * Grateful thanks are due to Shri D.V. Thakkar, Shri Manish Pandya, Shri S.V. Trivedi and Shri Rajendra Shah (all from Department of Economics and Statistics, Govt. of Gujarat) for providing basic source material and help in estimation of some variables. I am also thankful to Shri Akhilesh Kumar and Shri Saurabh Datta for valuable assistance.

2 Macroeconomic Framework for Development in Gujarat I. Introduction -Ravindra H. Dholakia I.I.M. Ahmedabad In the era of liberalization and globalisation, states have more freedom to decide on their development strategies, goals and policies. There is a growing competition among states to attract and encourage private sector activities into their territory. The competition could degenerate to offer several tax incentives and similar sops that may hurt the fiscal balance of all the states. A co-operative (or collusive!) solution has, therefore, emerged to avoid offering any such tax incentives and to agree on uniform floor rates of sales tax on different items. This would force the states to compete on tangible performance, cost & quality of infrastructural facilities, level of human development, living conditions and socio-cultural & business environment offered by each one. Development strategies, priorities and policies must address these aspects explicitly. Whatever development models and methodology for state level planning existed in the pre-liberalisation or pre-reform era are less relevant today. In those days, the state economies were not truly open, states did not have any assigned growth targets, there was little or no competition among states, private initiative was under constant curb and regulation, and resource reallocation was slow if not absent. Planners and statisticians did complain about data-gaps at the state level but the policy makers were hardly serious about the same (see, Committee on Regional Accounts, 1976 and National Statistical Commission, 2001). But now, the situation is fast changing. Policy makers at state level desperately need some macroeconomic framework and related data-set to formulate and monitor development programmes and policies to achieve the target that itself is the outcome of realistic assessment. The present paper makes a modest attempt in the direction of meeting this challenge in future by considering the case of Gujarat state. Gujarat is one of the leading states in the Indian Federation often considered as an engine of national economic growth and a frontline performer on various dimensions. Like most other economies of large magnitude, the development concerns in Gujarat 1

3 also pertain to the three dimensions: (i) economic growth aspects; (ii) human and social development aspect; and (iii) regional disparity aspect. Dimension of regional disparity in Gujarat has always been very important and remained at the root of state and sub-state level politics. There are 6 distinct geographical regions in Gujarat: (1) desert area of Kuchchh, (2) Saurashtra, (3) North Gujarat, (4) Eastern tribal belt, (5) Central Gujarat, and (6) South Gujarat. There are significant differences in the availability of natural resources, economic activity-mix, human and social development of population among these regions. Since Gujarat has decentralised multi-party democracy at each of the sub-state administrative levels of Districts, Talukas (or Blocks), Cities and Villages, the state government typically handles the aspect of regional disparity by providing special grants and development programmes to the backward or lagged areas in different sectors and spheres. In order to identify such problem areas, there are efforts from the central government to run some special area development programmes like Drought Prone Area Programme, Tribal Area Sub-Plan, Desert Area, Hilly Area, Costal Area, etc. and special efforts through committees of the state government. Thus, Gujarat government had the reports of Hathi Committee (1971) and I.G. Patel Committee (1983) to identify relatively lagged or backward talukas (blocks) in the state with heavy representation of academics on the committees. In spite of this, there have been practically no efforts to collect relevant production, consumption and employment oriented statistics and bridge the gaps in availability of data at the sub-state level. The approach of the state government in dealing with regional disparity has all along been to treat it more as an implementation issue than a design issue. This become evident when for two decades no effort went in for revising the list of identified backward talukas, and when recently the list was routinely and mechanistically revised internally by the bureaucrats without any inputs from the academics. The National Statistical Commission (2001) strongly recommended preparation and publication of the estimates of district domestic product (DDP) for all the districts in the state, but the state planning ministry and the Department of Economics and Statistics (DES) are yet not ready to carry out this important exercise. In a recent seminar of the Indian Association of Research in National Income and Wealth (IARNIW) 2

4 held at Jaipur in January 2004, this issue was discussed at length. Experts agreed that estimation of urban-rural income or productivity differential by sectors was a precondition for generating DDP for districts. Urban-rural gap is an important explanation for regional disparity (see, Dholakia & Dholakia, 1978). Although there were some individual (private) efforts made to estimate such sub-state level estimates in the distant past 1, no official or even individual estimate is available for the recent past. For our purposes, therefore, we have not considered explicitly the dimension of regional disparity firstly because ours is a macroeconomic framework whereas it is a micro issue; and secondly because the required type of data at sub-state level have yet not been collected for these units to integrate with the macroeconomic framework. In the next section, we examine the growth experience in various sectors over the last two decades so as to identify the growth potential of the state economy. In the third and fourth sections, we make an attempt to develop the elementary regional accounting framework and estimate the crucial macroeconomic aggregates and parameters for Gujarat. The fifth section discusses the implication of these estimates in terms of resource availability and the investment requirement to achieve the growth target derived from the assessment of the past performance. In order to identify the drivers of the economic growth in the state and hence derive broad strategy and policy implications, the sixth section develops macroeconomic model using the time series data over the past two decades. In the seventh section, we examine the human and social development aspects in Gujarat and see how we can integrate them with the macroeconomic model so as to derive strategic prescriptions. The eighth section concludes the paper by explicitly pointing out the need to develop official estimates of critical macro-aggregates and strategic policy interventions to achieve the development goals of the state. II. Economic Growth: Experience and Potential Economic growth has, in general, three connotations when referred to in professional or political circles. They are volume, efficiency and welfare. There is a wide spread 1 See for instance, Dholakia (1975), Dholakia (1976) and Kashyap (1984) for estimates of rural-urban incomes in Gujarat, district incomes in Gujarat and income for Ahmedabad respectively 3

5 consensus among the national income accountants and professionals that the aggregates of domestic product like GDP or GSDP by definition and design do not measure nor even reflect the welfare dimensions of the population. However, these aggregates measured in factor cost at constant prices do reflect most comprehensively the volume and efficiency aspects of total economic activities in the economy compared to any other known alternative measures. The Department of Economics & Statistics (DES) at the state level and Central Statistical Organisation (CSO) at the national level have been preparing the estimates of these aggregates at current and constant prices following almost identical methodology for a long time now. The estimates of GDP at national level and GSDP at the state level are available for 9 broad sector and 17 subsectors. The most recent base year is for all these estimates at constant prices. We, therefore, estimate the annual trend rates of growth during the last two decades in all these 17 sub-sectors in All-India and Gujarat with respectively GDP and GSDP at constant factor costs. Appendix Tables 1 and 2 report the detailed results and Table 1 summarises the comparison. Table 1: Comparison of Sectoral Trend Rates of Growth (% p.a.) Sector to to Gujarat India Gujarat India Agriculture Fishing Mining & Quarry Manufacturing Elect., Gas & Water Construction Trade & Hotels Transport & Communication Finance & Estates Public Administration Other Services Total GDP implies statistical insignificance Basic Source: DES (June 2003); and NAS, 2003 It is clear from Table 1 that Gujarat s performance was very poor compared to the nation during the decade of the eighties in all sectors except fisheries, electricity and 4

6 transport & communication. During the nineties, however, Gujarat improved its economic performance remarkably in almost all secondary and tertiary sectors except finance & real estates. While Gujarat has very strong performance in the manufacturing, electricity, construction, transport & communication, and services sector, its major weaknesses are in all the primary sectors and banking, insurance & real estate sectors. Statistical insignificance of the trend rate of growth in agriculture shows that it is a very heavily fluctuating activity in the state. In spite of such an indifferent performance of the primary sector, the total GSDP in Gujarat achieved 7% p.a. growth in real terms during the nineties. This growth is mainly on account of the secondary sectors, transport and services that have been the focus of the economic policy reforms in the nation. Thus, Gujarat not only benefits most by the liberalization and globalisation efforts initiated at the national level, but often acts like an engine of growth for the nation (see, Dholakia, 2003; and Ahluwalia, 2002) In order to examine the growth potential of the state in the immediate future, we need to consider a few similar efforts made for Gujarat. The Planning Commission (2002) has assigned the real growth target of 10.2% p.a. to Gujarat for the 10 th Plan period. Although it does not provide the precise methodological basis for its targets to different states, it has decomposed its target into the primary, secondary and tertiary sectors. Accordingly, Gujarat should achieve annual real growth rates of 4.3%, 12.23% and 10.44% respectively in the primary secondary and tertiary sectors. The Agro-Vision 2010, prepared by the Ministry of Agriculture (2001) in Gujarat, on the other hand, puts a very optimistic target of 6.8% p.a. for the real growth in agricultural sector in the state. However, by considering the trends in area, yield and productivity of 30 crops over last 30 years in Gujarat and other states in the country, it is possible to derive a plausibly optimistic growth target of about 5% p.a. for agriculture in Gujarat (see, Dholakia, 2003). For the remaining sectors, we can derive the growth potential by considering the past performance of the state during the last two decades. In order to identify the best episodes of growth in each sector and sub-sector of the economy, we should consider periods of four and ten consecutive years over the last two decades in the state. Such best growth episodes would reflect growth potential of the state in the medium term and long term respectively. Appendix Tables 3 and 4 report the annual growth rates over 4 5

7 and 10 consecutive years in different sectors in Gujarat over to , and Table 2 gives the potential growth rates as the maximum observed in each sub-sector of the state. Table 2: GSDP Share & Observed Maximum Growth in Consecutive Years (Gujarat) Sl. Max Growth Industry Group Share in GSDP No. (Potential %) Years 10 Year 1 Agri.& Allied Agriculture Forestry & Logging Fishing Mining & Quarrying Sub-total: Primary Manufacturing Registered Un-registered Elec., Gas & Water Construction Sub-total: Secondary Trade& Hotels Tran., Stor.& Comm Railways Other Transport Storage Communication Sub-total(6&7) Finance Sector Banking&Insurance Real Estate Comm. Services Public Adm Other Services Sub-total:Tertiary Total GSDP Basic Source: DES (2003): State Domestic Product, Gujarat State, ; GoG, June It is clear from Table 2 that Gujarat has achieved an overall GSDP growth of 11.9% p.a. in the medium term and 8.9% p.a. in the long-term. However, if we consider all the sectoral and sub-sectoral performance, we find that the potential could be much 6

8 higher for the state both in the medium term as well as in the long term. In order to get the idea about such an overall potential, we need to consider shares of the sub-sector in total GSDP at factor cost as reported in Table 2. Considering the sectoral shares in the latest year as the weights and replacing 5% p.a. growth potential of the agricultural sector, the overall growth potential of the state works out to 14.6% in the medium term and 9.4% in the long term. These targets are certainly very optimistic though falling on the outer border of feasibility. The targets of 11.9% p.a. and 8.9% p.a., on the other hand, are quite feasible considering that the state had actually achieved them in recent past. Any targets in between may be plausibly optimistic and need examination for consistency with other macroeconomic parameters of the state economy. III. Regional Accounts Estimates for Gujarat The Department of Economics & Statistics (DES) under the Planning Ministry in the state prepares and publishes regularly estimates of GSDP and NSDP at factor cost both at current and constant prices by sectors. Similarly, it also regularly publishes the Economic and Purpose Classification of State Budgets. Up to , it also prepared and published the estimates of Gross Domestic Capital Formation by the state public sector units. However, these are the only parts of the regional accounts prepared officially. In spite of the State Planning Department and the State Planning Board existing now for almost 40 years, the regional accounts are incomplete and absent in the state. Even individual efforts by researchers or any academic research projects on preparing regional accounts are absent. 2 This is not confined only to Gujarat but applies equally to all states in the country. 3 It is very surprising (if not shocking) to find that senior government officials and the professionals on the state planning boards could continue planning for the state all along these years without any idea about the broad 2 The field of input-output tables for states is an exception since considerable literature and several efforts at estimating I-O tables for states in India exist. However, almost all of these estimates suffer from the limitation about treatment of indirect taxes. For Gujarat, the I-O tables estimated by Alagh and Kashyap (1971), Kashyap (1976) and by Dholakia and Dholakia (1988) have been used in different planning models. 3 There are some ad hoc efforts to collect data on total investments in different states based on intensions, approvals and implementation (see, Tata Services Ltd., 2003). Their concepts are not consistent with national accounts statistics and their coverage is unknown. They can serve the limited purpose, but cannot serve to calculate the investment rate or the incremental capital output ratio, etc. 7

9 magnitude of regional accounts and any related aggregates. However, now the situation is fast changing. It is becoming imperative to have some broad idea about the aggregates involved in regional accounts to be able to frame right development strategies and policies. While it is best to generate detailed and direct estimates of all aggregates involved in the regional accounts, it is likely to be very time consuming and elaborate exercise. Generally, it will take years before the system is formally established, required surveys conducted and the estimates of relevant aggregates prepared and validated through formal procedures officially. In the meantime, however, we need to bridge the gap in the data availability by using all available information from various sources and making some bold but plausible assumptions. We make such a preliminary effort here to prepare the first cut estimates of aggregates in regional accounts of Gujarat state. The most well known fundamental identity of national accounts is: (1) Y = C + I + G + X M where Y is GDP at market prices; C is private final consumption expenditure (PFCE); I is investment expenditure or gross domestic capital formation (GDCF); G is the government s final consumption expenditure (GFCE); X is exports; and M is imports. This accounting identity applies equally to all regional economies defined in terms of geographical boundaries. Since states in India are completely integrated with the rest of the country sharing the same currency, it is relevant to distinguish between international trade flows and intra-national or domestic trade flows in the regional accounts. Thus, for Gujarat, we modify the above equation as: (2) Y = C + I + G + X F M F + X D + M D where the subscripts F and D represent respectively foreign and domestic flows. Let us now attempt to estimate each component of the identity (2) one by one. 3.1 GSDP at Market Prices: In collaborative effort with DES, estimates of GSDP at market prices for Gujarat were prepared from the estimates of GSDP at factor cost (see, Dholakia et.al.2002). For estimation, the indirect taxes and subsidies were divided into six components like: 1) net customs revenues; 2) net central excise; 3) indirect taxes of local bodies; 4) state indirect taxes; 5) state subsidies; and 6) state s share in centre s 8

10 subsidies. The study provides estimates of GSDP at market prices for the years to at current prices. However, the basic estimates of GSDP at factor cost for the years onwards have been subsequently revised. We should therefore revise the GSDP at market prices accordingly. Table 3 presents these revised estimates. It is important to observe from Table 3 that GSDP at market prices is almost 21% higher than the GSDP at factor cost in Gujarat in compared to only 9.9% at the national level. Year Table 3: Estimates of GSDP at Market Prices in Gujarat (Rs. in crore) GSDP at F.C. Net Net State State s State at customs central Octroi Indirect share in current revenues excise taxes * subsidy Central prices subsidy GSDP at market prices * state excise duty, taxes on vehicles, sales tax, entertainment tax, electricity duty, stamp duty and registration taxes on goods and passengers and tax on accommodation in hotels and lodges. Source: Dholakia et al. (2002); DES (June 2003); and SDP, Private Final Consumption Expenditure (PFCE): After the National Commission on Statistics (2001) explicitly recommending preparation of consumption estimates at the state level, there is some definite exploratory effort in this direction (see, Kar et al. 2004; CSO &NSSO, 2001, etc.). It is a general consensus that the commodity flow approach followed by the CSO for the national accounts is a better and more comprehensive method to estimate PFCE than the survey of direct consumption by NSSO, but that it is practically and conceptually very difficult to follow at the state level. CSO & NSSO (2001) attempt to reconcile these two sets of estimates at the national level by various items of consumption. Kar et al. (2004) also go into the details of the adjustments needed in the NSSO consumption survey data at a fairly disaggregated level. When it comes to making some ad hoc adjustment to restore consistency and comparability, however, it is better to operate at a reasonable degree of aggregation than a very detailed disaggregated level. We may, therefore, divide the total 9

11 consumption into food and non-food categories and apply the adjustment factors obtained from the study of CSO & NSSO (2001) to the consumption estimates in Gujarat from NSSO (2001) to generate per capita consumption expenditure comparable to the National Accounts Statistics (NAS). Then we can generate the estimate of total PFCE by considering the mid-year population of Gujarat for the year of the consumer survey, i.e Table 4 reports these estimates. Table 4: Estimates of Private Final Consumption in Gujarat, Items NSSO Estimates (in Per capita Monthly Rs.) Annual Estimates (in Rs.) Adjustment Factor for NAS NAS Comparable Estimates Items Rural Urban Total* (in (in. Rs, Rs.) Crores)** Food Non-Food Total Note: * Applying the proportion of urban and rural areas in Gujarat respectively at and as weights **Obtained by multiplying with estimates of mid-year population of crores Source: 1. NSS Report No. 461: Consumption of Some Important Commodities in India, ; Appendix Tables 2. DES (2003): SDP Gujarat, (June) 3. DES (2003): Statistical Abstract of Gujarat State, 2002 (March) 4. CSO & NSSO (2001): Report on Cross Validation Study of Estimates of PFCE Available from Household Survey and National Accounts 3.3 Government Final Consumption Expenditures (GFCE): As we have noted earlier, DES brings out annually the Economic and Purpose Classification of Budget of the state government. It contains an estimate of the final consumption expenditure by the state government. On specific request, the DES also made similar estimates for the local bodies in the state. However, in order to complete the estimation of GFCE consistent and comparable to the NAS, we need estimates of (i) central government units; (ii) consumption of fixed capital (CFC); and (iii) quasi-government bodies in the public sector. This is the major data gap existing at the state level and we may have to bridge it by taking ratios from the national level. Thus, we assume that the consumption of the administrative departments of state governments and local bodies as a proportion in the total consumption of all governments administrative departments remains the 10

12 same in Gujarat as in the nation for the given year. Based on this assumption, we generate the final consumption of administrative departments of different levels of government operating in Gujarat in different years. We may, then, assume the CFC and consumption by the quasi-government units in Gujarat bear the same proportion to such a total as in the nation. We can, thus, derive an estimate of GFCE for Gujarat consistent with the national estimate. Table 5 reports these estimates. We may note that GFCE in Gujarat in is about 5.4% of the national aggregate whereas the population of Gujarat is only 4.9%. Table 5: Public Sector Final Consumption Expenditure in Gujarat Administrative Departments (Rs. Crores) Year Total Multipliers Central Govt. State Local Central for CFC & Govt./ Total Total Final Govt. Bodies Govt. Quasi- PCFE Cons. Govt. Expn =3+4 6=(2*5)/ (1-2) 7= =7* Source: DES: Economic & Purpose Classification of Budgets, GoG (Annual); DES also for Local Bodies; For Central Govt. in Gujarat, the ratio for the nation for different years is applied to the total of (State Govt. + Local Bodies) obtained from NAS, (CSO) annual publication [EPWRF, (July 02), pp.108]; The Multipliers for Consumption of Fixed Capital and Quasi-Govt. bodies are obtained from EPWRF, (July, 02), pp.108 to obtain the total GFCE consistent with National Accounts 3.4 International Exports: The recent study by GITCO (2001) has very meticulously and carefully estimated the international exports originating from Gujarat. They have distinguished between the exports made from Gujarat ports and the exports originating from Gujarat and have accordingly estimated this very important aggregate for the Gujarat economy for the year They have conducted a huge sample survey covering 40% of the exporting units of the state. However, if we exclude a very 11

13 dominant but largely small and medium scale sector of Gems & Jewellery, in the remaining sectors, the sample covers almost 83% of the relevant units. Thus, the results are fairly reliable. Moreover, they have followed the market price approach as at the national level and hence the estimates are comparable to NAS. In order to derive the corresponding estimates of international exports originating in Gujarat for the year , we have assumed that their proportion in GSDP at factor cost (0.4481) has remained the same as in With this assumption, the international exports from Gujarat in work out to Rs crores. 3.5 International Imports: There are no studies attempting to estimate international imports made by units operating in Gujarat. We have estimated the customs duty collected from the units operating in Gujarat also on the basis of the principle of conservatism and not on actuals for estimating the GSDP at market prices. This is a serious gap in the availability of data at the state level. In order to overcome this gap, we may try to allocate the national imports to different states based on some relevant indicators. The indicators that we select for allocation must satisfy two criteria: (i) they must be highly significant determinants of the national aggregate we want to allocate; and (ii) they must be readily obtainable at the state level for the year under consideration, The following two indicators fully satisfy both these criteria: (1) income from registered manufacturing as a proportion in the total GDP at factor cost (i.e. RM/GDP fc ); and (2) exports as a proportion of GDP fc (i.e. X F /GDP fc ). At the national level, these two indicators explain more than 99.5% of variation in Imports to GDP fc ratio over the period : (3) (M F /GDP fc ) = 0.148(RM/GDP fc ) (X F /GDP fc ) t-values (8.72) (27.28) : Adj. R 2 = (M F /GSDP fc )Guj = (0.2286) (0.4481) = (in ) International Imports in Gujarat in = Rs crores. 12

14 3.6 Gross Domestic Capital Formation (GDCF): This is a major lacuna in the regional accounts in most of the states. There are only three to four states in the country who attempt to estimate GDCF in their economies. There are several states who prepare the estimates of capital formation by the government sector only. Gujarat is one of them. However, it leaves out a large part of the economy uncovered for estimating the overall GDCF in the state. There are 3 major components of GDCF: (i) public sector investment (PSI); (ii) household physical investment (HHPI); and (iii) private corporate sector investment (PCSI). Data on all these three components are regularly available at the national level. Again in this case, we may follow the method of allocating the national totals using indicators that satisfy the two criteria mentioned in section 3.5 above. The following indicators fully satisfy these criteria: (1) growth of GDP over last two years (G GDP.2 ), and capital expenditure as proportion of GDP fc (i.e. CE/GDP fc ) for public sector investment (i.e. PSI/GDP fc ); (2) income from construction sector as a proportion of total GDP fc (i.e. Const/GDP fc ) for the household physical investment (i.e. HHPI/ GDP fc ); and (3) HHPI/ GDP fc, income from electricity & gas as a proportion of GDP fc (i.e. EGW/GDP fc ), income from unregistered manufacturing sector as a proportion of GDP fc (i.e. URM/ GDP fc ), income from agriculture & animal husbandry as a proportion of GDP fc (i.e. AA/GDP fc), and G GDP.2 for the private corporate sector investment (i.e. PCSI/ GDP fc ). These indicators show a very high degree of explanatory power at the national level over the period : (4) (PSI/ GDP fc ) = (G GDP.2 ) (CE/ GDP fc ) t-values: (11.05) (12.05) :Adj. R 2 =0.996 (PSI/GSDPfc)Guj. = (1.1802) (0.0468) = PSI in Gujarat in = Rs.9234 crores. (5) (HHPI/ GSDPfc) = (Const/ GSDPfc) t-values: (33.64) :Adj.R 2 = (HHPI/ GSDPfc)Guj. = (0.0594) = HHPI in Gujarat in = Rs crores. 13

15 (6) (PCSI/GDP fc )= (HHPI/GDP fc )+3.861(EGW/GDP fc )+0.466(URM/GDP fc ) t-values: (-5.41) (2.68) (2.12) 0.879(AA/GDP fc )+0.199(G GDP.2 ) (-3.62) (4.23) :Adj.R 2 = (PCSI/GDP fc ) Guj. = PCSI in Gujarat in = Rs crores. (7) GDCF = PSI + HHPI + PCSI = = Rs crores This estimate implies that Gujarat s share is 7.9% in the national aggregate. 3.7 Domestic Export-Import Gap (X D -M D ): Based on the above estimates for the state, we can derive an estimate of the domestic export-import gap as the residual. Thus, considering our estimates derived in sub-sector 3.1 to 3.6 above, we get (8) GSDP mp = C + I + G + X F M F + (X D -M D ) i.e = (X D -M D ) (X D -M D ) = + Rs crores; and (9) M D = X D The central sales tax (CST) on most items sold to other states from Gujarat of the value net of transport. As a result, we can get an indirect estimate of the domestic exports of Gujarat from the CST collection of Rs crores in the state during This would imply a minimum domestic export of Rs crores from Gujarat. It may be more on account of exemptions or evasions but cannot be less. Thus, the minimum domestic imports in Gujarat would be 5378 crores. 14

16 IV. Estimate of Saving Rate in Gujarat The implications of the estimates of the regional accounts presented above are startling and far reaching. It shows that Gujarat s share in the nation in GDP at market prices is 6.65%, in private final consumption expenditure is 5.24%, in the gross domestic capital formation is 7.89%, in the government final consumption is 5.38%, in international exports is 20.87%, and in international imports is 20.11%. Thus, the general impression that Gujarat is a trading state gets overwhelming support from these estimates. These estimates also show that Gujarat is not a major consuming state, nor does it have a larger government sector. Its investment share is higher than its income implying that it is more investment oriented compared to the nation. Moreover, our estimation methodology for investment and domestic exports & imports makes it clear that the errors in the two estimates are in the opposite direction and mutually cancelling. It should, therefore, be possible to generate some crude estimate of savings in Gujarat. The second most famous accounting identity involves savings (S), investment (I), trade gap (X-M), and the government deficit from its tax revenues: (10) S = I + (XF-MF) + (XD-MD) + (G + Subsidies + Other Current Transfers Govt. Taxes from the State.) It is clear that once we get estimates of the transfer payments and taxes by government sector ascribable to Gujarat, we can get the estimate of savings in the state. Tables 6 and 7 provide these estimates by broad items of government transfers and taxes respectively with the sources and method of estimation. From these estimates, we get savings in Gujarat as: (11) S = (-3255) ( ) = Rs crores This implies that the saving rate (S/GDP mp ) in Gujarat is 38.94%. Moreover, in spite of very high tax collection from Gujarat, on the whole the government sector runs a marginal deficit in terms of its consumption and transfer payments in the state. This 15

17 goes against the general impression that Gujarat is a net contributor to the national kitty. The impression is right though in terms of savings, not the government sector. Table 6: Transfer Payments of Government Sector in Gujarat, (Rs. Crores) Level of Item Tr. Payment Source Govt. to Guj Grants 1154 Budget Documents 1. Central Govt. 2. State Govt. 3. Local Bodies Subsidies 1077 Dholakia et. Al. (2002) Interest 5237 Misc. 375 Economic Survey, provides an estimate of interest on internal liabilities of Central Govt. (p.31). We assume that Gujarat s share would be the same as in GDP at factor cost Economic Survey, provides an estimate of Rs crores for its total current transfers (p/35). Deducting grants, subsidies and interest from the current transfers, about Rs.12,500 crores remain for various other items like scholarships, grants to educational and other institution, etc. We have assumed that Gujarat gets about 3% of this amount. Sub-Total 7843 Subsidies 1550 DES (July, 2002) EPCB Others (Net) Total Transfers on Current A/c Sub-Total 7386 Transfer Payments -- Likely to be negligible (Net) Interest payment by GoG to the Central Govt. should be netted out. Thus, Rs.1894 crores (see, Finance and accounts, , p.97) is deducted. Similarly imputed losses of Rs.1387 crores of irrigation schemes are also deducted since it represents a contra entry item; DES (July, 2002) EPCB The savings in Gujarat contribute 11.64% of the national aggregate and the saving rate is as high as 39%. Thus, Gujarat is a high saving society, but invests only a fraction of what it saves in the state. It invests in a big way outside the state and also receives returns. Its savings rate is very comparable to those prevailing in China, Korea, `Malaysia, Japan, Singapore, Honk Kong, etc. thus, problems in Gujarat is not so much 16

18 to increase savings but to retain its savings and utilise productively giving high returns to the investors. Thus, although foreign direct investment and capital may be very important on margin, the major challenge before the policy makers and planners is to retain the savings in the state by attracting even the domestic investors to invest in Gujarat. The problem needs further data and analysis of the components of the savings particularly in terms of the household financial savings, corporate sector s savings and public sector savings. This is because savers in Gujarat are also financially very active, investing in a big way in the share market, mutual funds, company deposits and the property market. Most of these savings by very nature does not have to remain within the state. Although most of the policy makers and professionals have the right impression about the linkages of the health of stock market and the consumers wellbeing in Gujarat, there are no formal estimates of these linkages. For want of any reliable data, we cannot examine or analyse these issues now, but the need to develop these estimates through DES cannot be over emphasised. Table 7: Government Taxes Ascribable to Gujarat, (Rs. Crores) Level of Tax Item Govt. Amount Source Central Indirect Taxes Dholakia et.al. (2002) Govt. Direct Taxes 2390 DES (Feb. 2003) SER, State Indirect Taxes 7823 DES (July, 2002) EPCB Govt. Direct Taxes 234 DES (July, 2002) EPCB 3.Local Indirect Taxes 872 Dholakia et.al. (2002) Bodies Direct Taxes -- Likely to be negligible Total Taxes from Gujarat Eco. Survey , p.41 (25.01% of Tax Rev.) Misc. Receipts of Govt. 2.24% of Taxes (EPWRF, p.32) 623 Admin. Depts. Total

19 Another major implication of all these estimates presented here is in term of the overall investment rate in the state economy. The total investment (i.e. GDCF) in the state is Rs crores out of the GSDP at market prices of Rs crores in The investment rate works out to 26.93%, it means that overall export surplus of Gujarat is about 12% of GSDP. This includes both international as well as domestic export surplus. Before we derive any further implication, it is important to pause and look into our methodology of estimation and the nature of the year for which we have estimated the aggregates. The investment or capital formation estimates for Gujarat are essentially dependent on allocation of the corresponding aggregate at the national level are themselves fluctuating rather than stable from year to year. Moreover, the year is a peculiar year in the sense that it was almost a normal and average year for the nation clocking the growth in real terms at 6.2% in GDP at factor cost, but for Gujarat it was a bad year when GSDP actually declined by 1.2% on account of very bad monsoon in the state. We should, therefore, expect a depressed investment rate during the year in Gujarat. If we take a long term average rate of investment in Gujarat, it would be around 28% to 29% of GSDP at market prices. This is because the incremental capital-output ratio (ICOR) for the nation is around 4 during the period and the trend rate of growth in Gujarat over the last decade was 7% (see Table 1). Since the share of the secondary sector in general and of registered manufacturing sector in particular is significantly higher in Gujarat than the nation, we should expect a marginally higher ICOR in Gujarat than in India. Considering all this, we may still get an estimate of net export surplus of about 10% in Gujarat. It means that Gujarat s own resources can generate additional real growth of about 2.5% p.a. assuming the same ICOR as in the past. However, the national ICOR is likely to be at 3.6 as assumed or targeted in the Tenth Five Year Plan (2001). If Gujarat succeeds to bring down its ICOR from around 4 to even 3.8, its saving rate of 38.94% would imply 10.25% growth p.a. in real terms. If Gujarat s ICOR is at 3.6, its growth rate could be 10.8% p.a. However, this involves a Herculean, or more appropriately, a Bhagirath effort at retaining or diverting all the savings generated in Gujarat to get productively invested in Gujarat. Alternatively, attract equal amount of inflows of investment (i.e.10%) from outside the state including foreign direct investment in the state. 18

20 V. Alternative Growth Targets and Macroeconomic Implications We have seen earlier that based on the performance in the recent past, Gujarat has the potential to achieve annual growth rate up to 14.6% in the medium term and 9.4% in the long term. However, if we consider the saving rate in the state and also assume the investment rate of the same order, Gujarat can achieve a long term growth of about 10.8% p.a. when we take the optimistic ICOR of 3.6 as targeted in the 10 th Five Year Plan (2001). We, therefore, need to consider alternative growth targets and the implied macroeconomic aggregates to get an idea about the magnitude of effort required and the nature of the policy options to consider. We have to note, however, that critical macroeconomic parameter estimates are simply not available for the state since estimation of regional accounts is incomplete and inadequate. Even at the national level, some parameters like the factor shares are not available in the directly usable form, but enough material exists to allow construction of the required estimates (see, Dholakia, Bakul; 2001). Fortunately, such parameters are not likely to substantially vary across states. Hence, we can use the national parameters for Gujarat without risk of high error. The famous Harrod-Domar growth identity provides the link between the growth target, ICOR and the required investment rate: (12) (I/Y) = ICOR (G Y *) where Y is GSDP and G* is the growth target. In order to achieve this growth target coupled with the employment growth target, we need several necessarily implied targets achieved. They include the overall growth of capital stock, the risk free interest on capital, internal rate of return (IRR), the rate of technical progress, the wage income as a proportion in the new investment, etc. All these can be worked out consistently through the well-known neo-classical growth theory framework. Thus, we have an aggregate production function with output (Y) in factors-capital (K) and labour (L), and time (t): (13) Y = f(k,l, t) ; then dy/dt = f K (dk/dt) + f L (dl/dt) + f t ; and hence (14) G Y = R K * G K + R L * G L + r which is the famous neo-classical growth accounting equation, where R K and R L are relative factor shares; G K and G L are growth rates of capital and labour; and r is the residual also known as rate of technical progress. Similarly (13) also implies 19

21 (15) (dy/dk) = f K + (f L *dl/dk) + (f t /(dk/dt)) where f K is the risk free interest; (f L *dl/dk) in the wage income in new investment; f K + (f t /(dk/dt)) is IRR; and (dy/dk) is the reciprocal of the ICOR. If we assume equilibrium conditions, ICOR as well as relative factor shares become constant and it is possible to work out all implications of growth targets with given values of ICOR, R K and R L. Table 8 provides alternative growth targets and their macroeconomic implications. Table 8: Alternative Growth Targets and Implied Macroeconomic Parameters Macro With G L = 3% p.a. With G L = 3.5% p.a. Parameters G Y = 9% G Y = 10% G Y = 12% G Y = 9% G Y = 10% G Y = 12% Growth of Capital (G K ) 9% 10% 12% 9% 10% 12% Investment Rate (I/Y) 32.4% 36% 43.2% 32.4% 36% 43.2% Return on Capital 21.8% 22.4% 23.3% 20.8% 21.5% 22.5% (IRR) Wage Income as % in New 6% 5.4% 4.5% 7% 6.3% 5.3% Investment (f L dl/dk) Rate of Technical Progress (r) 4.8% 5.6% 7.1% 4.5% 5.3% 6.7% Note: Assumptions are: ICOR = 3.6; risk free interest (f k ) = 6.9%; relative shares of labour and capital are 0.65 and 0.25 respectively We can see from the table that high growth of employment coupled with high growth of income would require very high investment rate in the state. 43.2% compared to the current rate of about 28.5% requires huge inflow of capital that can come only in the form of FDI. It is not impossible, but calls forth bold decisions and policy changes to ensure an IRR on the new projects well above 22%. This again is achievable by singularly focusing on innovations, research & development, improving the quality of products, better management & organisation, skills improvement, and very fundamental changes in the structures of economic activities shifting away from low productivity traditional ones to high value, high productivity modern activities. This calls forth very bold decisions on allowing entrepreneurial flexibility in resource shift and allocation, which is not possible without allowing free exit and removing all technical, administrative and economic barriers in such movement. 20

22 VI. Drivers of Growth An Econometric Model In order to achieve specific growth target, it is important to identify certain drivers of growth in the system. Very distant past experience may not be of much use in such estimation of the current relationships. The econometric exercise is, therefore, inherently limited to considering relatively recent time series data. We can consider the last two decades as the relevant time span for our purpose. The number of observations are, therefore, limited to 22, from to There are hardly any quarterly or monthly series on relevant variables available at regional or sub-regional level. Any question of using panel data or sub-annual data simply does not arise. Moreover, the constraint on the number of observations also imposes restrictions on the size of the model in terms of number of exogenous variables. We must recognise and appreciate that with all such constraints, the econometrically estimable and meaningful model will have to evolve slowly. It would need careful scrutiny, interpretation and validation. What we are now discussing can only qualify as exploratory and tentative effort. We can begin by identifying a few most relevant growth oriented and targeted variables, called endogenous variables. We need to determine or target their values in future. Each of these variables depends on several of those variables where either the government exercise some control or outside factors determine their values putting constraint on our postulated relationships. These are the exogenous or pre-determined variables. Our drivers of growth would belong to this category. Based on intuitively appealing causal links, we can postulate the structural form of the model. Table 9 and 10 provide respectively the description of the endogenous and the exogenous variables used in the model. All the nominally measured variables are in real terms after correcting for the inflation through the GSDP deflator. The income variables are, however, available at constant base period prices and do not require any further deflation. We consider 8 endogenous and 14 exogenous variables in the model. Table 9: The Endogenous Variables of the Model Endogenous Variables (8) Variable Notation Agriculture (Agri.) and Fishery Y 1 Manufacturing (Mfg.) Y 2 Trade and Transport (TT) Y 3 Financial, Administrative & Other Services (Service) Y 4 Government Total Non Interest Expenditure (GITNIE) Y 5 Government Own Tax Revenue (GOTR) Y 6 State Income (GSDP) Y 7 Modern Inputs in Agriculture (MAI) Y 8 21

23 Table 10: The Exogenous Variables of the Model Endogenous Variables (14) Forestry (Forest ) X 1 Man-days Lost (Man DL/MDL) X 8 Government Expenditure on Human Government Non Tax Revenue (GNTR) X 9 Capital (GEHK) X 2 Government Expenditure on Physical Real Estate, Ownership of dwelling & Business Capital (GEPK) X 3 activity (RE) X 10 Rainfall (Rain) X 4 Transfer from the Centre (TFC) X 11 Storage & Communication (Storcom) X 5 Electricity, Gas & Water (EGW) X 12 Construction (Const) X 6 Mining & quarrying (MQ) X 13 Wage Rate (WR) X 7 Capital-Output Ratio (COR) X 14 Eqn. No. 1 Agri Dependent Variable 2 Manufacturing 3 TT 4 Service 5 6 Govt. Total Non Int Exp Govt. Own Tax Revenue 7 GSDP 8 Modern Inputs Govt. Total Non Int Expenditure Table 11: Structure of the Model and Test of Identification Independent Variable k M (K-k) (M-m) Inference Storage & Communication, Electricity Water & Gas, Rainfall, Modern Inputs, Government Expenditure on Physical Capital Agri, TT, Forestry, EWG, Storage & Communication, Real Estate, Govt. Exp. On PK Agri, Manufacturing, EWG, Storage & Communication, Construction, Real Estate, Govt Exp on PK, Manufacturing, TT, EWG, Real Estate, Govt Exp on HK, Storage & Communication, Construction Govt. Own Tax Revenue, Govt. Non Tax Revenue, Transfer from Centre, GSDP Manufacturing, Construction, EWG Agri, Manufacturing, TT, Services Government Expenditure on PK, EWG, Storage & Communication, Rainfall Identity Over Identified Exactly Identified Over Over Over Over Over Over Govt Exp on HK, Govt Exp on PK Notes: T=Total number of variables included in the model =8+14=22 M=Number of endogenous variables included in the model =8 K=Number of exogenous variables included in the model =14 m=number of endogenous variables in the given equation k=number of exogenous variables in the given equation N=Number of Observations =22